Thermal management is a growing challenge for modern aircraft. Onboard Thermal Management Systems (TMS) must content with both external heating and internal heat loads. To meet these challenges, TMS rely heavily on Compact Heat Exchangers (CHEs). This thesis explores the design and performance of gyroid-based CHEs. Enabled by advances in Additive Manufacturing (AM), gyroid structures--classified as Triply Periodic Minimal Surfaces (TPMS)--offer high surface area-to-volume ratios and a continuous, interwoven geometry well-suited for dual-stream devices. Using Computational Fluid Dynamics (CFD), this work models such devices across a wide range of operating conditions, providing insight into the viability of gyroid-based designs for TMS.
To establish a foundation, a validated CFD framework is developed using a two-dimensional representation of a traditional wavy-fin plate-fin CHE to assess the impact of various numerical modelling assumptions. It is found that representative hydraulic and thermal performance of a full-scale wavy-fin HE can be recovered, with significantly reduced computational cost, by simulating periodic flow and heat transfer in a 2D wavy-fin test section, as validated against experimental correlations from the literature across a wide range of Reynolds numbers (20 < Re < 5500).
A practical model for evaluating the performance of gyroid-based CHEs was developed by simulating a three-dimensional gyroid test section defined by periodic length (L) and offset parameter (R), which control unit cell size and wall thickness, respectively. The model captures both steady and unsteady effects and is validated against larger-domain simulations, transient solutions, and published data. Based on this framework, two parametric studies were performed, yielding original empirical correlations for the Fanning friction factor (f) and Chilton-Colburn J-factor (j) over 5 < L < 40 mm, 0 < R < 0.6 mm, and 20 < Re < 8000. These correlations quantify sensitivity to geometric variation and offer a predictive tool for integrating gyroid-based CHEs into TMS.
A conjugate cross-flow heat exchanger model was developed to more realistically represent a gyroid-based CHE. This approach captures key physical effects such as transverse mixing and inter-stream thermal interaction, which were absent in the previous approach.
To investigate novel solutions for flow maldistribution--a core challenge in CHE design--this thesis evaluates the gyroid’s ability to self-correct under asymmetry conditions and benchmarks it against a sinusoidal-channel HE. The gyroid’s interconnected structure redistributes non-uniform inflow within 2–3 unit cells, improving outlet uniformity by 16.5% and maintaining thermal-hydraulic performance within 4% across 150 < Re < 8000. In contrast, the sinusoidal geometry, characterised by disconnected flow paths, exhibited significant performance variability under the same conditions. This previously unreported behaviour highlights the gyroid’s suitability for systems where flow maldistribution is unavoidable and active correction impractical.
Finally, the potential of multi-directional geometric grading in gyroid-based heat exchangers was explored through two warp-based strategies, contracting and expanding the core center. While grading altered local flow and pressure due to changes in density and path tortuosity, thermal performance remained largely unchanged (< 0.6% variation). Both strategies increased pressure drop (7.3% contracted, 4.6% expanded), reducing overall thermo-hydraulic efficiency. Although these strategies did not improve performance, they demonstrate the potential of geometric grading as a tool for targeted flow control in gyroid-based CHEs.
Overall, this thesis advances the understanding of gyroid-based CHEs. Empirical correlations for performance are developed in relation to the geometric parameters. In addition, gyroid-specific behaviours relevant to real-world challenges have been investigated. Collectively, these findings offer valuable insight into the viability of gyroid-based designs for integration into TMS.
